JP2014010139A - State monitoring system for dynamic facility and its method and its program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the state monitoring system of a dynamic facility capable of performing highly accurate and highly efficient state monitoring while performing processing with simple configurations when a plurality of mechanical equipment whose drive frequency is close are arranged and driven under a narrow space or an environment which is low in rigidity.SOLUTION: The state monitoring system of a dynamic facility includes: n pieces of sensors; a frequency measurement part for acquiring n pieces of composite vibration waveform signals; a fast Fourier transformation processing part for performing the fast Fourier transformation processing of the n pieces of composite vibration waveform signals to acquire a spectrum signal; a waveform separation processing part for separating only vibration wave components generated from the most adjacent vibration source among n pieces of vibration sources; an inverse fast Fourier transformation processing part for performing the inverse fast Fourier transformation processing of the spectrum signal to acquire a vibration waveform signal from the most adjacent vibration source; a residual calculation part for calculating a difference between the vibration waveform signal in a normal time and the acquired vibration waveform signal; and an evaluation part for determining that the vibration source is not normal.

Description

  The present invention relates to a dynamic equipment state monitoring system and method capable of separating synthesized vibration waveforms generated from a plurality of vibration sources, and in particular, utilizing intensity change due to distance attenuation of vibration waveform or directivity change of acoustic waveform. In addition, the present invention relates to a dynamic equipment state monitoring system, a method thereof, and a program thereof that can identify a dynamic equipment causing a problem by separating a synthesized vibration waveform.

In dynamic equipment equipped with rotating devices such as rotating motors and gears or rolling bearings, vibrations that are different from normal may occur due to misalignment of the rotating shaft, imbalance or bearing damage, or scratches or wear on gear parts. It has been known.
Such vibrations vary in frequency and intensity (amplitude) depending on the power supply frequency, the number of rotations of the rotating shaft, and where the scratches and wear occur, and differ in the characteristics of the vibration frequency detected in the normal state, but are detected in the normal state. Because it is measured as a composite vibration waveform mixed with vibration frequency, it is difficult to detect the abnormality that has occurred unless only the vibration waveform based on the abnormality is separated and analyzed. It was difficult.
Also, it is common that a plurality of dynamic facilities are actually installed on the same floor, such as production factories, chemical plants or power plants, and are driven at close drive frequencies. There is also a need for state monitoring.
Under such circumstances, conventionally, techniques for separating a specific waveform from a synthesized waveform have been developed.

For example, Patent Document 1 discloses a waveform separation program that can separate each waveform when there is a plurality of synthesized waveforms that are not known even with one basic waveform under the name “waveform separation program and method”. Is disclosed.
In this waveform separation program, one of the combined waveforms is arbitrarily multiplied to obtain a difference waveform from the other synthesized waveform, one of the obtained difference waveforms is assumed as a basic waveform, and this basic candidate waveform is arbitrarily multiplied and separated. Assuming a waveform, another separation candidate waveform obtained in the same manner as this separation candidate waveform is synthesized and assumed as a synthesis waveform, and the synthesis candidate waveform is compared with one synthesis waveform to obtain the first error, and the other is obtained. The second error is obtained by comparison with the synthesized waveform, and a plurality of candidate synthesis waveforms are obtained by using each of the means, and the first error and the second error are further obtained. The combination candidate waveform having the smallest sum of the errors and the combination of the basic candidate waveforms forming the combination candidate waveform is identified, and the separation candidate waveform forming the combination candidate waveform is output.

  Patent Document 2 discloses a technique for removing pulsating noise from a musical tone signal under the name of “noise removal method, noise removal apparatus and program”. In this technique, a synthesized waveform is obtained by shifting a portion that does not include the pulsating noise by a certain amount of time that includes the pulsating noise.

JP 2007-93555 A JP 2005-309464 A

  However, in the invention disclosed in Patent Document 1, the processing is performed in the time domain, and the waveform separation is performed on the principle of minimizing the sum of errors of the synthesis candidate waveform. In this method, there is a problem that it is impossible to completely remove the frequency component to be separated (to zero). In addition, in the invention disclosed in Patent Document 1, it is impossible to accurately reproduce or leave a frequency component having a statistically random property generated when a dynamic facility is normal. There has been a problem that it is difficult to obtain information necessary for performing.

  Further, in the invention disclosed in Patent Document 2, since the vibration waveform based on the abnormality is deleted rather than separated, the superimposed normal vibration waveform is also deleted, and a highly accurate separation method and There was a problem that could not be said.

  The present invention has been made in response to such a conventional situation, and has a simple configuration when a plurality of mechanical equipments having close drive frequencies are arranged and driven in a narrow space or a low rigidity environment. It is an object of the present invention to provide a dynamic equipment state monitoring system, a method thereof, and a program thereof that are capable of highly accurate and highly efficient state monitoring while performing processing.

In order to achieve the above object, in the dynamic equipment state monitoring system according to the first aspect of the present invention, in an environment where n (n is an integer of 2 or more) vibration sources exist, the n vibration sources. A dynamic facility state monitoring system capable of separating a combined vibration waveform generated from the n number of sensors capable of receiving the combined vibration wave generated from the n number of vibration sources and received by the n number of sensors. A frequency measurement unit that obtains n synthesized vibration waveform signals by processing the synthesized vibration waves, and fast Fourier transform (hereinafter referred to as fast Fourier transform) of the n synthesized vibration waveform signals obtained by the frequency measurement unit. (Sometimes referred to as FFT) to obtain a spectral signal by processing, and for the n spectral signals obtained by the fast Fourier transform processing unit, take a difference in intensity for each spectrum, Only the spectrum signal of the spectrum having a positive difference with respect to any of the spectrum signals is left, and the spectrum signal of the other spectrum is set to 0, and is generated from the nearest vibration source among the n vibration sources. A waveform separation processing unit that separates only the vibration wave component to be processed, and an inverse fast Fourier transform processing that obtains a vibration waveform signal from the nearest vibration source by performing an inverse fast Fourier transform process on the spectrum signal separated by the waveform separation processing unit And a residual for calculating the difference between the vibration waveform signal at normal time from the closest vibration source measured in advance and the vibration waveform signal from the closest vibration source obtained from the inverse fast Fourier transform processing unit. The calculation unit and evaluation for comparing the difference with a desired threshold value and determining that the closest vibration source is not normal when the deviation is larger than the threshold value And having a, the.
In the dynamic equipment state monitoring system having the above configuration, the waveform separation processing unit takes the difference between n spectrum signals for each spectrum and a spectrum signal other than the spectrum signal itself, Even if only the spectrum signal of the spectrum with the positive difference is left, the spectrum signal of the other spectrum is set to 0, so that the spectrum signal attenuated by the distance is excluded, and the synthesized vibration waveform signal from the sensor is excluded. To extract as a vibration waveform signal from the vibration source closest to. Also, the difference between the vibration waveform signal of the closest vibration source and the normal vibration waveform signal at the residual output unit is calculated, and the evaluation unit compares the difference with a desired threshold value and the difference is large. The vibration waveform signal extracted by the previous waveform separation processing unit is evaluated as not normal, and thus the closest vibration source is evaluated as not normal.
Note that n sensors do not necessarily have to exist. When the dynamic equipment is operating in a steady state, for example, one sensor is moved to each of n vibration sources. That is, it may be used for reception n times. In other words, the n sensors mean “total” n sensors. The same applies hereinafter.

A dynamic equipment state monitoring system according to a second aspect of the present invention is the dynamic equipment state monitoring system according to the first aspect, wherein the n synthetic vibration waveform signals obtained by the frequency measuring unit are envelope-processed. An envelope processing unit is provided, and the fast Fourier transform processing unit inputs a post-envelope synthesized vibration waveform signal generated by the envelope processing unit, and performs a fast Fourier transform process to obtain a spectrum signal.
In the state monitoring system for dynamic equipment having the above-described configuration, in addition to the operation of the first aspect, the envelope processing unit takes the absolute value of the raw composite vibration waveform and smoothes it with a low-pass filter. Have Therefore, the post-envelope synthesized vibration waveform signal generated by the envelope processing unit does not have a fine shocking free vibration component, and misalignment of the rotating shaft of the dynamic equipment, imbalance or bearing damage, or gear parts In other words, a combined vibration waveform signal is generated that gently changes in accordance with the period of abnormal vibration caused by scratches and wear.

The dynamic facility state monitoring system according to claim 3 is the dynamic facility state monitoring system according to claim 1 or 2, wherein the synthesized vibration waves generated from the n vibration sources are combined. Instead of n sensors that are sound waves and can receive the synthesized vibration wave, a microphone that can measure the synthesized sound wave generated from the n vibration sources by changing the direction n times with a single directivity. And the frequency measuring unit processes each of the synthesized sound waves received by the microphone to obtain n number of synthesized sound waveform signals as n number of synthesized vibration waveform signals.
In the dynamic equipment state monitoring system having the above configuration, the direction is changed n times by using one microphone having unidirectionality instead of the n sensors in claims 1 and 2. Thus, this single microphone acts to receive n synthetic sound waveform signals in the same manner as the n sensors. Other functions are the same as those of the first or second aspect of the invention.

According to a fourth aspect of the present invention, there is provided a dynamic facility state monitoring method for separating synthesized vibration waveforms generated from n vibration sources in an environment where n (n is an integer of 2 or more) vibration sources exist. A method for monitoring the state of a possible dynamic facility, wherein a reception step of receiving a composite vibration wave generated from the n vibration sources using n sensors, and a composite vibration wave received in the reception step A frequency measurement process for obtaining n synthesized vibration waveform signals by processing each of the signals, and a fast Fourier transform process for obtaining a spectrum signal by performing a fast Fourier transform process on the n synthesized vibration waveform signals obtained in the frequency measurement process; The spectrum of the spectrum in which the difference in intensity is obtained for each spectrum with respect to the n spectrum signals obtained in the fast Fourier transform process, and the difference is positive with respect to any other spectrum signal. The waveform separation processing step for separating only the vibration wave component generated from the nearest vibration source among the n vibration sources with the spectral signal of the other spectrum being set to 0, leaving only the signal, and the waveform separation An inverse fast Fourier transform processing step for obtaining a vibration waveform signal from the nearest vibration source by performing an inverse fast Fourier transform process on the spectrum signal separated in the processing step, and a normal time from the nearest vibration source measured in advance A residual calculation step of calculating a difference between the vibration waveform signal and the vibration waveform signal from the nearest vibration source obtained in the inverse fast Fourier transform processing step, and comparing the difference with a desired threshold value An evaluation step of determining that the closest vibration source is not normal when the deviation is larger than a threshold value.
In the dynamic equipment state monitoring method having the above-described configuration, the dynamic equipment state monitoring system according to claim 1 is regarded as a method invention, and the operation thereof is the state of the dynamic equipment according to claim 1. It is the same as the monitoring system.

According to a fifth aspect of the present invention, there is provided the dynamic equipment state monitoring method according to the fourth aspect, wherein the n synthetic vibration waveform signals obtained in the frequency measurement step are enveloped. An envelope processing step, wherein the fast Fourier transform processing step receives the post-envelope synthesized vibration waveform signal generated in the envelope processing step and performs a fast Fourier transform process to obtain a spectrum signal.
In the dynamic equipment state monitoring method having the above-described configuration, the dynamic equipment state monitoring system according to claim 2 is regarded as a method invention, and the operation thereof is the state of the dynamic equipment according to claim 2. It is the same as the monitoring system.

The dynamic equipment state monitoring method according to claim 6 is the dynamic equipment state monitoring method according to claim 4 or 5, wherein the synthesized vibration waves generated from the n vibration sources are synthesized. Instead of the n sensors, the receiving step measures a synthetic sound wave generated from the n vibration sources by changing the direction n times with a single directivity using a microphone. In the receiving step, the frequency measuring step processes the synthesized sound waves received by the microphone to obtain n number of synthesized sound waveform signals as n number of synthesized vibration waveform signals.
In the dynamic equipment state monitoring method having the above-described configuration, the dynamic equipment state monitoring system according to claim 3 is regarded as a method invention, and the action thereof is the state of the dynamic equipment according to claim 3. It is the same as the monitoring system.

The dynamic facility state monitoring program according to the invention of claim 7 is a dynamic facility state monitoring program executed by a computer to separate a composite vibration waveform generated from n vibration sources, A reception step for receiving the composite vibration wave generated from the n vibration sources using the n sensors, and a composite vibration wave received in the reception step are respectively processed to obtain n composite vibration waveform signals. A frequency measurement step, a fast Fourier transform processing step for obtaining a spectrum signal by performing a fast Fourier transform process on the n synthesized vibration waveform signals obtained in the frequency measurement step, and n obtained in the fast Fourier transform process step For each spectrum signal, the difference in intensity is taken for each spectrum, and only the spectrum signal of the spectrum for which the difference is positive for any other spectrum signal remains. Thus, the spectrum signal of the other spectrum is set to 0, and the waveform separation processing step for separating only the vibration wave component generated from the nearest vibration source among the n vibration sources is separated by the waveform separation processing step. An inverse fast Fourier transform processing step for obtaining a vibration waveform signal from the closest vibration source by performing an inverse fast Fourier transform process on the spectrum signal, and a normal vibration waveform signal from the closest vibration source measured in advance, A residual calculation step of calculating a difference between the vibration waveform signals from the closest vibration source obtained in the inverse fast Fourier transform processing step, and comparing the difference with a desired threshold value from the threshold value; And an evaluation step of determining that the closest vibration source is not normal when the deviation is large.
In the dynamic equipment state monitoring program having the above-described configuration, the dynamic equipment state monitoring method described in claim 4 is regarded as a program invention, and the action thereof is the state of the dynamic equipment according to claim 4. This is the same as the monitoring method.

The dynamic facility state monitoring program according to the invention of claim 8 includes an envelope processing step of performing envelope processing on the n synthesized vibration waveform signals obtained in the frequency measurement step, and the fast Fourier transform processing step includes: A post-envelope synthesized vibration waveform signal generated in the envelope processing step is input, and a fast Fourier transform process is performed to obtain a spectrum signal.
In the dynamic equipment state monitoring program having the above configuration, the dynamic equipment state monitoring method described in claim 5 is regarded as a program invention, and the operation thereof is the state of the dynamic equipment according to claim 5. This is the same as the monitoring method.
It has the action.

According to a ninth aspect of the present invention, there is provided the dynamic facility state monitoring program, wherein the synthesized vibration wave generated from the n number of vibration sources is a synthesized sound wave, and the reception step is performed in place of the n number of sensors. A receiving step of measuring a synthesized sound wave generated from the n number of vibration sources using a microphone by changing the direction n times with one directivity, wherein the frequency measuring step is a synthesis received by the microphone. Each of the sound waves is processed to obtain n synthetic sound waveform signals as n synthetic vibration waveform signals.
In the dynamic facility state monitoring program having the above-described configuration, the dynamic facility state monitoring method described in claim 6 is regarded as a program invention, and the operation thereof is the state of the dynamic facility according to claim 6. This is the same as the monitoring method.

According to the dynamic facility state monitoring system of the first aspect of the present invention, the synthesized vibration waveform signal received by the sensor and obtained by the frequency measuring unit is converted into a spectrum signal by the fast Fourier transform processing unit. Each time, the waveform separation processing unit can extract the vibration waveform signal of the closest vibration source from the combined vibration waveform signals of the n vibration sources. Therefore, the abnormal vibration source can be easily identified with high accuracy.
Further, by providing n sensors as many as the number of vibration sources, it is possible to detect whether or not abnormal vibration has occurred from any of the vibration sources, and the waveform separation processing unit can detect from the n sensors. As described above, the received n synthesized vibration waveform signals are compared for each spectrum, and a difference is obtained. A spectrum signal of a spectrum in which all the differences are positive is left, and the spectrum of the spectrum obtained from another sensor is obtained. By setting the spectrum signal to 0, only the vibration waveform received by the sensor closest to the vibration source can be separated efficiently and with high accuracy.
Therefore, it is possible to efficiently and highly accurately monitor the state of the dynamic facility.
Furthermore, if the status of dynamic equipment can be monitored efficiently and with high accuracy, the maintenance policy based on time-based maintenance can increase the maintenance interval, and thus maintenance costs. There are advantages such as being able to reduce the number of spare parts and the number of spare parts. Furthermore, since sudden stoppage of facilities and the like can be prevented in advance, damage can be minimized.

  According to the dynamic facility state monitoring system of the second aspect of the present invention, since the envelope processing unit can be provided to smooth the synthesized vibration waveform, noise can be eliminated. Therefore, it is possible to more easily extract the vibration waveform accompanying the abnormal vibration included in the combined vibration waveform.

  According to the dynamic facility state monitoring system of the third aspect of the present invention, it is possible to monitor the state by measuring the sound generated from the dynamic facility. In addition, the number of sensors can be reduced by changing the direction n times using a unidirectional microphone, so that sound waves can be measured more efficiently and at the same time the cost can be reduced. Is also possible.

  Since the dynamic equipment state monitoring method according to claim 4 of the present invention is an invention which captures the dynamic equipment state monitoring system according to claim 1 as a method invention, the effect thereof is the invention according to claim 1. It is the same.

  Since the dynamic equipment state monitoring method according to claim 5 of the present invention is an invention in which the dynamic equipment state monitoring system according to claim 2 is regarded as a method invention, the effect thereof is the invention according to claim 2. It is the same.

  Since the dynamic equipment state monitoring method according to claim 6 of the present invention is an invention which grasps the dynamic equipment state monitoring system according to claim 3 as a method invention, the effect thereof is the invention according to claim 3. It is the same.

  Since the dynamic equipment state monitoring program according to claim 7 of the present invention is an invention that captures the dynamic equipment state monitoring method according to claim 4 as a software invention, the effect thereof is the invention according to claim 4. It is the same.

  Since the dynamic equipment state monitoring program according to claim 8 of the present invention is an invention that captures the dynamic equipment state monitoring method according to claim 5 as a software invention, the effect thereof is the invention according to claim 5. It is the same.

  Since the dynamic facility state monitoring program according to claim 9 of the present invention is an invention that captures the dynamic facility state monitoring method according to claim 6 as a software invention, the effect thereof is the invention according to claim 6. It is the same.

1 is a system configuration diagram of a dynamic facility state monitoring system according to an embodiment of the present invention. It is a flowchart of the signal processing in the state monitoring system of the dynamic installation which concerns on this Embodiment. It is a photograph which shows the structure and sensor arrangement | positioning of the dynamic equipment of the monitoring object in the test conducted using the state monitoring system of the dynamic equipment of the Example of this invention. (A) is a graph which shows the synthetic | combination vibration waveform signal measured with the sensor and vibration frequency measurement part of the main shaft side in an Example, (b) is the synthetic vibration measured with the sensor and vibration frequency measurement part of the driven shaft side. It is a graph which shows a waveform signal. (A) is a graph showing a combined vibration waveform spectrum signal obtained by performing a fast Fourier transform process on the combined vibration waveform signal measured by the spindle-side sensor and vibration frequency measurement unit in the embodiment, and (b) These are the graphs which show the synthetic | combination vibration waveform spectrum signal which carried out the fast Fourier transform process of the synthetic | combination vibration waveform signal measured by the sensor and vibration frequency measurement part in the driven shaft side in an Example. (C) is the graph which extracted the range of 0 to 70 Hz of the frequency of the synthetic vibration waveform spectrum signal shown to (a), (d) is 0 to 70 Hz of the frequency of the synthetic vibration waveform spectrum signal shown to (b). It is the graph which extracted the range. It is a graph which shows together the separated vibration waveform spectrum signal which carried out waveform separation processing of the synthetic vibration waveform spectrum signal by the side of a principal axis in an example, and the spectrum signal which carried out waveform separation processing. It is a graph which shows together the separated vibration waveform spectrum signal which carried out the waveform separation process of the synthetic | combination vibration waveform spectrum signal by the side of a driven shaft in an Example, and the spectrum signal by which the waveform separation process was carried out. It is a graph which compares and shows the isolation | separation vibration waveform signal after carrying out the inverse fast Fourier-transform process of the synthetic | combination vibration waveform signal by the side of a spindle in an Example, and a separation vibration waveform spectrum signal. (A) is a graph showing a differential vibration waveform signal obtained by taking a difference between the combined vibration waveform signal on the spindle side shown in FIG. 8 and the separated vibration waveform signal after the inverse fast Fourier transform processing is performed on the separated vibration waveform spectrum signal. (B) is a graph which shows the difference vibration waveform signal which took the difference of the separated vibration waveform signal after carrying out the inverse fast Fourier-transform process of the synthetic vibration waveform signal of a driven shaft side, and a separated vibration waveform spectrum signal similarly. (A) is a graph in which the differential vibration waveform signal shown in FIG. 9 (a) is subjected to fast Fourier transform, and the power spectrum is displayed with the frequency as the horizontal axis, and (b) is the differential vibration shown in FIG. 9 (b). It is the graph which carried out the fast Fourier transform of the waveform signal, and displayed the power spectrum on the frequency as a horizontal axis.

A dynamic facility state monitoring system according to an embodiment of the present invention will be described in detail with reference to FIG.
FIG. 1 is a system configuration diagram of a dynamic facility state monitoring system according to an embodiment of the present invention, and FIG. 2 is a flowchart of signal processing in the dynamic facility state monitoring system according to the present embodiment.

As shown in FIG. 1, a dynamic equipment state monitoring system 1 according to an embodiment of the present invention includes a sensor 2, a vibration frequency measuring unit 3, an envelope processing unit 4, a fast Fourier transform processing unit 5, and a waveform separation processing unit. 6, an inverse fast Fourier transform processing unit 7, a residual calculation unit 8, an evaluation unit 9, and an output unit 10, and a combined vibration waveform database 11 for storing measurement signals, processed signals, and threshold values for evaluation , A spectrum database 12, a separated wave waveform database 13, and an evaluation database 14. The dynamic facility state monitoring system 1 composed of these components has a plurality of sensors 2 but is not complicated in arithmetic processing and is simply configured.
Hereinafter, these system configurations will be described with reference to FIGS. 1 and 2.

In FIG. 1, a plurality of sensors 2 are provided and are arranged in the vicinity of a dynamic facility in which a plurality of electric rotating machines or the like as vibration sources are installed to detect a combined vibration waveform from the plurality of vibration sources (FIG. 2). Step S1). The vibration frequency measurement unit 3 processes the combined vibration waves from the plurality of sensors 2 to obtain the combined vibration waveform signal 15 as many as the number of sensors 2 (the number is n in the present embodiment) (see FIG. Step S1) in step 2. Since each sensor 2 corresponds to a vibration source, it is desirable that the number of the sensors 2 matches the number of vibration sources. This is because this method separates the synthesized waveform with high accuracy by using the intensity change due to the distance attenuation of the vibration waveform. Since it is necessary to measure the vibration of each vibration source in the vicinity, the number of sensors 2 corresponding to the number of each vibration source is necessary. However, when the dynamic equipment to be monitored is operating in a steady state, it is possible to move a single sensor to the vicinity of each vibration source and shift the measurement time, and the same effect can be obtained. it can. That is, the number of sensors 2 may be “total number”. In this embodiment, it is assumed that the number of vibration sources included in the dynamic facility is n.
The synthesized vibration waveform signal 15 generated by the vibration frequency measuring unit 3 is stored in the synthesized vibration waveform database 11 so as to be readable by the vibration frequency measuring unit 3 (step S1 in FIG. 2).

The number of sensors 2 generated by the vibration frequency measurement unit 3, that is, n composite vibration waveform signals 15, are directly read from the vibration frequency measurement unit 3 or input from the composite vibration waveform database 11 by the envelope processing unit 4. Then, smoothing is performed (step S2 in FIG. 2).
The envelope processing unit 4 is equipped with a so-called low-pass filter, which eliminates shocking high-frequency components and accompanies abnormal vibrations due to misalignment or imbalance of the rotating shaft of dynamic equipment or bearing damage, or scratches or wear on gear parts. It is possible to generate a combined vibration waveform signal that gently changes in accordance with the period of abnormal vibration. The envelope processing unit 4 is not necessarily required, but it is desirable to provide the envelope processing unit 4 in order to improve the accuracy of the evaluation of the presence or absence of abnormality by the dynamic facility state monitoring system 1. The post-envelope synthetic wave waveform signal 16 generated by the envelope processing unit 4 is readable and stored in the synthetic vibration waveform database 11 by the envelope processing unit 4 (step S2 in FIG. 2).
The n post-envelope synthesized wave waveform signals 16 generated by the envelope processing unit 4 or, if the envelope processing unit 4 is not provided, the n synthesized vibration waveform signals generated by the vibration frequency measuring unit 3 are The data is read from the envelope processing unit 4 or the vibration frequency measuring unit 3 directly or from the combined vibration waveform database 11 by the fast Fourier transform processing unit 5 and input to the fast Fourier transform processing unit 5 (step S3 in FIG. 2).

The fast Fourier transform processing unit 5 performs fast Fourier transform processing on the n synthesized vibration waveform signals to generate n synthesized vibration waveform spectrum signals 17 (step S3a in FIG. 2). Further, the fast Fourier transform processing unit 5 calculates an effective value (root mean square value) of the combined vibration waveform spectrum signal 17 to obtain a combined vibration waveform power spectrum signal 24 (step S3b in FIG. 2). Since this combined vibration waveform power spectrum signal 24 includes the intensity of each spectrum as a signal, it is suitable for analyzing an abnormality having a unique frequency generated from any of a plurality of vibration sources. it is conceivable that. The combined vibration waveform spectrum signal 17 and the combined vibration waveform power spectrum signal 24 generated by the fast Fourier transform processing unit 5 are readable and stored in the spectrum database 12 by the fast Fourier transform processing unit 5 (step S3a in FIG. 2). , S3b).
In addition, as shown in FIG. 2, steps S1 to S3a and S3b may be repeatedly performed to obtain an average value to improve the SN ratio.

The waveform separation processing unit 6 includes each of the n synthesized vibration waveform power spectrum signals 24 generated by the fast Fourier transform processing unit 5 and the other (n−1) synthesized vibration waveform power spectrum signals 24 other than itself. A difference in intensity is taken for each spectrum (step S4 in FIG. 2). Accordingly, n (n-1) differences are obtained for each spectrum, but the frequency of the power spectrum signal having a spectrum that is positive with respect to any other power spectrum signal is detected. When the frequency is detected by the waveform separation processing unit 6, the waveform separation processing unit 6 reads the synthesized vibration waveform spectrum signal 17 directly from the fast Fourier transform processing unit 5 or from the spectrum database 12, and the spectrum signal related to the frequency is The process of leaving the spectrum signals of other frequencies as 0 (zero) is executed. Therefore, this waveform separation processing is also called zero processing. In the waveform separation processing unit 6, by using the composite vibration waveform power spectrum signal 24 at the time of zero processing, it is possible to improve the accuracy when taking the difference.
By this waveform separation processing, the spectrum signal in the spectrum of the combined vibration waveform signal detected by the sensor 2 having the maximum spectrum signal in the n synthesized vibration waveform spectrum signals 17 is left, and other sensors 2 are left. In the synthetic vibration waveform spectrum signal 17 at, the spectrum signal in the spectrum can be zero. That is, it is possible to obtain information on which sensor 2 detects the vibration waveform related to the spectrum.
Moreover, for example, even when a spectrum signal related to a vibration waveform generated from another vibration source is included in the vicinity of this spectrum, it is detected as the strongest signal by the sensor 2 closest to the other vibration source. For the synthesized vibration waveform spectrum signal 17 generated from this vibration source, the spectrum signal in the spectrum in the vicinity thereof becomes 0 as described above, so that the spectrum signals from other vibration sources in the vicinity of the spectrum are excluded. Will be.
Therefore, it can be separated from other vibration waveform components as a spectrum signal of the vibration waveform from the vibration source closest to the sensor 2. In the present application, the vibration waveform spectrum signal thus separated is referred to as a separated vibration waveform spectrum signal 18.
By performing such processing in all spectral bands, a unique vibration waveform from the vibration source closest to the sensor 2 is detected with the characteristic spectral signal shown in the specific sensor 2. Can be understood.
The waveform separation processing unit 6 detects a specific vibration waveform from the vibration source closest to each sensor 2 with respect to the combined vibration waveform spectrum signal 17 obtained from the combined vibration waveform received by each sensor 2. Is possible.
The separated vibration waveform spectrum signal 18 processed by the waveform separation processing unit 6 is readable and stored in the spectrum database 12 by the waveform separation processing unit 6 (step S4 in FIG. 2).

The inverse fast Fourier transform processing unit 7 inputs a spectrum signal from which only the vibration wave component from the closest vibration source is separated by the waveform separation processing unit 6, that is, the separated vibration waveform spectrum signal 18, and inputs this spectrum signal to the inverse fast Fourier transform. To convert. Therefore, the inverse fast Fourier transform processing unit 7 can obtain the vibration waveform signal from the vibration source closest to each sensor 2, that is, the separated vibration waveform signal 19 (step S5 in FIG. 2).
The separated vibration waveform spectrum signal 18 is read directly from the waveform separation processing unit 6 by the inverse fast Fourier transform processing unit 7 or read from the spectrum database 12 and processed by the inverse fast Fourier transform processing unit 7.
The separated vibration waveform signal 19 generated by the inverse fast Fourier transform processing unit 7 is stored in the separated wave waveform database 13 so as to be readable by the inverse fast Fourier transform processing unit 7 (step S5 in FIG. 2).

The residual calculation unit 8 includes a separated vibration waveform signal 19 detected from the vibration source closest to each of the n sensors 2, and a normal wave waveform signal at the vibration source, that is, a normal vibration waveform signal 20. Is calculated (step S6 in FIG. 2). The separated vibration waveform signal 19 is read directly from the inverse fast Fourier transform processing unit 7 or from the separated wave waveform database 13, and the normal-time vibration waveform signal 20 is stored in the separated wave waveform database 13 in advance. The difference is read and calculated, and a differential vibration waveform signal 21 is generated. Thus, if there is a difference between the normal vibration waveform signal 20 and the separated vibration waveform signal 19, it is considered that some abnormality may have occurred in the vibration source closest to the sensor 2.
The differential vibration waveform signal 21 generated by the residual calculation unit 8 is stored in the separated wave waveform database 13 so as to be readable by the residual calculation unit 8 (step S6 in FIG. 2).

The evaluation unit 9 compares the differential vibration waveform signal 21 obtained by the residual calculation unit 8 with the differential signal threshold value 22 stored in the evaluation database 14 in advance. It is evaluated that an abnormality has occurred in the vibration source (step S7 in FIG. 2). The evaluation result 23 by the evaluation unit 9 includes the difference between the difference vibration waveform signal 21 and the difference signal threshold value 22 and the presence or absence of abnormality, and this evaluation result 23 is stored in the evaluation database 14 so as to be readable by the evaluation unit 9. (Step S7 in FIG. 2).
Whether or not the deviation is large depends on the difference from the difference signal threshold 22, and when the difference vibration waveform signal 21 is larger than the difference signal threshold 22, it is determined that the deviation is large. It is. The difference signal threshold 22 is set according to the dynamic equipment to be monitored, the location and type of abnormality, the importance of the purpose of monitoring, and the like. Good.

  The output unit 10 is generated by the signals processed by the components of the dynamic equipment state monitoring system 1 from the vibration frequency measuring unit 3 to the evaluation unit 9 described so far, and the signals, values, and processes used for the processing. The output signal can be output (step S8 in FIG. 2). Specifically, a display device such as a display device such as a CRT, liquid crystal, plasma, or organic EL, a tablet, or a printer device, or a transmitter such as a transmitter for transmission to an external device may be considered. Of course, it may be an interface for output for transmission to an external device. At the time of output, the output unit 10 may directly read from the components (sensor 2, vibration frequency measurement unit 3 to evaluation unit 9) of the dynamic equipment state monitoring system 1 described in FIG. You may read from each database (the synthetic | combination vibration waveform database 11, the spectrum database 12, the isolation | separation wave waveform database 13, the evaluation database 14) of the state monitoring system 1 of a dynamic installation.

In the present embodiment, the state monitoring for evaluating the presence or absence of abnormality in the vibration source by measuring the vibration wave generated in the vibration source constituting the dynamic equipment has been described. Sound waves are also included, and the vibration wave in the above description is considered as sound waves. In that case, instead of n sensors 2, there may be provided a microphone that can measure the synthesized sound wave generated from n vibration sources n times by changing the direction n times with unidirectionality. Good. Specific examples of such microphones include microphones having a single directivity and microphones equipped with a parabolic sound collector in order to have a single directivity. When using such a microphone, it is possible to monitor the state by receiving a sound wave while changing the direction of the microphone and estimating the wave propagation direction. For example, when measurement is performed by changing the direction n times, a combined vibration waveform signal 15 is obtained for each direction, and the vibration source closest to that direction is detected. Therefore, it is important to have unidirectionality in order to improve accuracy.
By receiving sound waves while changing the direction of a microphone having unidirectionality, the installation is easier than installing a plurality of sensors, the configuration is simplified, and the cost can be reduced. Moreover, since the propagation direction of the sound wave is understood to some extent by the direction change, the composite vibration waveform signal 15 can be measured efficiently with a small number of measurements.
Further, the vibration frequency measuring unit 3 is established as a sound frequency measuring unit, and the signal to be measured is a combined vibration waveform signal 15 as a synthesized sound waveform signal, and a synthetic wave waveform signal 16 after envelope processing is a synthesized sound waveform signal after synthesis. The vibration waveform spectrum signal 17 is a synthetic sound waveform spectrum signal, the synthetic vibration waveform power spectrum signal 24 is a synthetic sound waveform power spectrum signal, the separated vibration waveform spectrum signal 18 is a separated sound waveform spectrum signal, and the separated vibration waveform signal 19 is a separated sound waveform signal. The normal vibration waveform signal 20 is also established as a normal sound waveform signal, and the differential vibration waveform signal 21 is also established as a differential sound waveform signal.

Further, in this embodiment, the state monitoring system 1 of the dynamic equipment has been described mainly using FIG. 1 as well as FIG. 2, but as another embodiment, a method and a method for monitoring the state of the dynamic equipment are provided. There is a dynamic facility state monitoring program that is regarded as a program for executing a computer.
The contents of these other embodiments are as shown in the flow of steps (processes) described in FIG.
That is, by executing the steps from step S1 to step S8, it becomes an embodiment relating to the dynamic equipment state monitoring method, and furthermore, by executing these steps using a computer, the dynamic equipment state monitoring program and It becomes. Although specifically described below, the description of the processing contents in each step is omitted because it has already been implemented when the dynamic facility state monitoring system 1 is described.

  Specifically, as shown in FIG. 2, in step S <b> 1 as synthetic vibration waveform measurement, the synthetic vibration waveform signal 15 is obtained by using the hardware of the sensor 2 and the vibration frequency measurement unit 3 to obtain the synthetic vibration waveform database 11. To be readable.

  In step S2, the envelope processing unit 4 reads the composite vibration waveform signal 15 directly from the vibration frequency measurement unit 3 or from the composite vibration waveform database 11, performs envelope processing, and obtains a composite wave waveform signal 16 after envelope processing, Similarly, it is stored in the synthetic vibration waveform database 11 so as to be readable. However, as described above, this envelope processing step is not necessarily required, but it is desirable to provide it in order to improve the accuracy of the presence / absence evaluation by the dynamic facility state monitoring method or its program.

In step S3a, the fast Fourier transform processing unit 5 reads the composite wave waveform signal 16 after envelope processing directly from the envelope processing unit 4 or from the synthetic vibration waveform database 11, performs fast Fourier transform processing, and performs the composite vibration waveform spectrum signal 17. And stored in the spectrum database 12 in a readable manner.
In step S <b> 3 b, the fast Fourier transform processing unit 5 calculates the effective value (root mean square value) of the combined vibration waveform spectrum signal 17 to obtain the combined vibration waveform power spectrum signal 24 and stores it in the spectrum database 12 so that it can be read out. To do.
As described above, step S1 to step S3 may be repeated a desired number of times to obtain an average value and improve the SN ratio.

  In step S <b> 4, the waveform separation processing unit 6 reads the synthesized vibration waveform power spectrum signal 24 directly from the fast Fourier transform processing unit 5 or from the spectrum database 12 and analyzes the frequency to be subjected to zero processing and the remaining frequency. Thereafter, the waveform separation processing unit 6 reads the combined vibration waveform spectrum signal 17 and performs the waveform separation processing for performing the zero processing for the frequency for which the previous zero processing is performed to obtain the separated vibration waveform spectrum signal 18 to obtain the spectrum database. 12 is readable and stored.

  In step S5, the inverse fast Fourier transform processing unit 7 reads the separated vibration waveform spectrum signal 18 directly from the waveform separation processing unit 6 or from the spectrum database 12, and performs an inverse fast Fourier transform process to obtain a separated vibration waveform signal 19. And stored in the separated wave waveform database 13 so as to be readable.

  In step S 6, the residual calculation unit 8 reads the separated vibration waveform signal 19 directly from the inverse fast Fourier transform processing unit 7 or from the separated wave waveform database 13, and further, the normal-time vibration waveform signal 20 from the separated wave waveform database 13. And the difference is calculated to obtain a differential vibration waveform signal 21 and stored in the separated wave waveform database 13 so as to be readable.

  In step S7, the evaluation unit 9 reads the differential vibration waveform signal 21 directly from the residual calculation unit 8 or from the separated wave waveform database 13, and further reads the differential signal threshold value 22 from the evaluation database 14 and compares them. If the differential vibration waveform signal 21 exceeds the differential signal threshold value 22, it is evaluated that an abnormality has occurred in the vibration source of the closest dynamic facility measured by the sensor 2 and must be exceeded. For example, the evaluation result 23 relating to the presence / absence of an abnormality to be evaluated as having no abnormality is stored in the evaluation database 14 in a readable manner.

  In step S <b> 8, the output unit 10 reads out and outputs signals, values, or evaluation results that are considered to be output from the components of the dynamic facility state monitoring system 1 or the database of the dynamic facility state monitoring system 1. Is.

  As described above, in the dynamic facility state monitoring system or method and program according to the present embodiment, the vibration waveform received by the sensor closest to the vibration source in the dynamic facility while having a simple configuration. Can be separated efficiently and with high accuracy. As a result, when multiple dynamic equipment with similar drive frequencies are placed and driven in a narrow space or in a low-rigidity environment, the status of these dynamic equipment is monitored efficiently and with high accuracy. can do. Furthermore, by performing envelope processing, the synthesized vibration waveform can be smoothed and noise can be eliminated. Therefore, it is possible to more easily extract the vibration waveform accompanying the abnormal vibration included in the composite vibration waveform, and the state monitoring system can be further enhanced.

  A dynamic facility state monitoring system 1 according to an embodiment of the present invention will be described with reference to FIGS. 3 to 10.

FIG. 3 is a photograph showing the configuration of the dynamic equipment to be monitored and the arrangement of sensors in the test conducted using the dynamic equipment status monitoring system of the embodiment of the present invention. In FIG. 3, the dynamic equipment includes a main shaft 30 connected to a power source (not shown) and an electric motor driven by the power source, and the main shaft 30 is provided with a rotor 34 and a main shaft side gear 32. A driven shaft side gear 33 that meshes with the side gear 32 is provided on the driven shaft 31.
This rotor 34 has an artificial imbalance so as to cause unbalance due to the load weight and local wear of the driven shaft side gear 33 at one place (for two gears).
A main shaft side sensor 35 is installed on the main shaft 30 and a driven shaft side sensor 36 is installed on the driven shaft 31 to receive vibration waveforms.
In the dynamic equipment configured as described above, the frequency of the electric motor is 60 Hz, and the rotation speed N of the main shaft 30 is 500 (rpm). Therefore, the rotation frequency f1 of the main shaft 30 is approximately f1 = N / 60 = 500/60. It becomes 8.33 Hz, and as its harmonics, the second order component is 16.88 Hz, the third order component is 24.93 Hz, and the fourth order component is 33.24 Hz. Further, since the number of teeth Z1 of the main shaft side gear 32 is 40 and the number of teeth Z2 of the driven shaft side gear 33 is 43, the rotational frequency f2 of the driven shaft 31 is f2 = f1 · Z1 / Z2 = about 7. As a harmonic, the second order component is 15.44 Hz, the third order component is 23.19 Hz, and the fourth order component is 30.64 Hz. Therefore, abnormal components are included at these frequencies due to the above-mentioned artificial troubles.
In this embodiment, the vibration sampling frequency is 5 kHz, and the cutoff frequency is 400 Hz.
Although not shown in FIG. 3, the vibration waveforms received by the main shaft side sensor 35 and the driven shaft side sensor 36 have the same configuration as that of the dynamic facility state monitoring system 1 according to the above-described embodiment. The analysis processing is possible as a vibration waveform generated in relation to the main shaft 30 and a vibration waveform generated in relation to the driven shaft 31. Accordingly, also in the present embodiment, the components already described in the embodiment will be described using the same reference numerals, and description of the configuration will be omitted. However, in this embodiment, since the vibration waveform is received using the two sensors of the main shaft side sensor 35 and the driven shaft side sensor 36, the vibration waveform received on the main shaft side and the vibration waveform received on the driven shaft side are distinguished. In the signals related to each of them, a subordinate character “a” is attached to the main shaft side and “b” is attached to the driven shaft side.
In the present embodiment, the rotor 34 on the main shaft 30 side is artificially imbalanced, and the driven shaft side gear 33 on the driven shaft 31 side is artificially subjected to local wear, causing abnormalities on the respective shafts. Is imitating the state.

In the present embodiment, whether or not the current dynamic equipment state monitoring system 1 can actually carry out the dynamic equipment state monitoring with high accuracy has been tested by the configuration shown in FIG. 3, this will be described.
4A is a graph showing the combined vibration waveform signal 15a measured by the main shaft side sensor 35 and the vibration frequency measuring unit 3 in the embodiment, and FIG. 4B is a graph showing the driven shaft side sensor 36 and the vibration frequency measuring unit 3. It is a graph which shows the synthetic | combination vibration waveform signal 15b measured by (1). In each graph, the horizontal axis represents time (s: second), and the vertical axis represents intensity (unit: V (volt)). The bolt with this strength indicates the output of the vibration frequency measuring unit 3.
In the graph shown in FIG. 4, the main shaft side sensor 35 is based on an unbalanced rotor 34 as well as a normal vibration waveform related to 60 Hz of the main power source and a vibration waveform related to the main shaft side gear 32. In addition to the abnormal vibration waveform, an abnormal vibration waveform associated with local wear of the driven shaft side gear 33 is superimposed and received.
On the other hand, the driven shaft side sensor 36 not only has a vibration waveform related to 60 Hz of the main power supply, but also an abnormal vibration waveform due to local wear of the driven shaft side gear 33, and an abnormal vibration based on the unbalance of the rotor 34. A thing related to a vibration waveform is also superimposed and received.
However, since it is difficult to determine which vibration waveform is generated from which normal or abnormal configuration by looking at these graphs (a) and (b), this implementation is performed. The effect was tested using the dynamic facility state monitoring system 1 according to the embodiment.
The combined vibration waveform signals 15a and 15b shown in FIGS. 4 (a) and 4 (b) are subjected to envelope processing by the envelope processing unit 4 and synthesized after envelope processing (not shown), respectively, as described in the embodiment. Wave waveform signals 16a and 16b are generated, and these are subjected to fast Fourier transform processing by the fast Fourier transform processing unit 5 to generate combined vibration waveform spectrum signals 17a and 17b.
In this embodiment, the envelope processing is performed. However, as long as sufficient accuracy can be obtained without performing the envelope processing, the envelope processing is not particularly required as described above.

5A shows a combined vibration waveform spectrum signal 17a obtained by performing a fast Fourier transform process on the combined vibration waveform signal 15a measured by the spindle sensor 35 and the vibration frequency measuring unit 3 in the embodiment. (B) is a combined vibration waveform spectrum signal obtained by subjecting the combined vibration waveform signal 15b measured by the driven shaft side sensor 36 and the vibration frequency measuring unit 3 in the embodiment to a fast Fourier transform processing by the fast Fourier transform processing unit 5. It is a graph which shows 17b. (C) is the graph which extracted the range (location shown with a rectangle) of 0 to 70 Hz of the frequency of the synthetic | combination vibration waveform spectrum signal 17a shown to (a), (d) is the synthetic | combination vibration waveform shown to (b). It is the graph which extracted the range (location shown with a rectangle) of 0 to 70 Hz of the frequency of the spectrum signal 17b. In either case, the horizontal axis represents frequency (Hz), and the vertical axis represents the FFT value indicating the intensity (unit: V (volt)).
The reason why the low frequency part is extracted and shown or the signal of this part is processed in this way is that the present invention is intended for state monitoring for structural abnormality. Therefore, also in the present embodiment, vibrations based on mechanical abnormalities such as rotor unbalance and gear local wear are actually generated artificially. 5A to 5D, the peak occurs at the frequency of 60 Hz because of the frequency of the drive power supply.

FIG. 6 also shows a signal obtained by performing zero processing on the separated vibration waveform spectrum signal 18a that has been subjected to zero processing by the waveform separation processing unit 6 with respect to the synthetic vibration waveform spectrum signal 17a thus shown. FIG. 7 shows the separated vibration waveform spectrum signal 18b similarly to the vibration waveform spectrum signal 17b. 6 and 7, the horizontal axis represents frequency (Hz), and the vertical axis represents the FFT value indicating the intensity (the unit is V (volt)).
In FIG. 6, the portion indicated by the symbol X and the portion indicated by the symbol Y in FIG. 7 is the place where the zero processing has been performed. That is, the portion where the zero processing is performed is indicated by a dotted line. Specifically, in FIG. 6, when the vibration waveform generated in the driven shaft 31 is received, the driven shaft side sensor 36 is higher than the main shaft side sensor 35, so that the main shaft side sensor 35 generates vibration in the driven shaft 31. The signal for the waveform is zeroed. That is, in terms of frequency, they are 7.75 Hz, 15.44 Hz, 23.19 Hz, and 30.64 Hz indicated by the symbol X including harmonics.
Similarly, in FIG. 7, the frequencies indicated by the symbol Y are 8.33 Hz, 16.88 Hz, 24.93 Hz, and 33.24 Hz. 60 Hz is zero-processed in both FIGS. 6 and 7 because it is clear that this is the power supply frequency.

FIG. 8 is a graph showing a comparison between the combined vibration waveform signal on the main shaft 30 side and the separated vibration waveform signal after the inverse fast Fourier transform processing of the separated vibration waveform spectrum signal in the embodiment. The horizontal axis of FIG. 8 is time (s: second), and the vertical axis is intensity (unit: V (volt)). Further, the part indicated by the solid line is the combined vibration waveform signal, and the part indicated by the dotted line is the separated vibration waveform signal after the inverse fast Fourier transform processing of the separated vibration waveform spectrum signal.
In FIG. 8, the combined vibration waveform signal 15a received by the main shaft side sensor 35 indicated by the solid line includes a vibration waveform based on the abnormality of the dynamic equipment generated on the driven shaft 31, but is indicated by the dotted line. In the separated vibration waveform spectrum signal 18 a zero-processed by the waveform separation processing unit 6, the vibration waveform based on the abnormality of the dynamic equipment generated in the driven shaft 31 is excluded. Therefore, only the vibration waveform generated in the main shaft 30 remains, and an abnormality occurring in the vicinity of the main shaft 30 can be detected.

FIG. 9A shows a differential vibration waveform signal obtained by taking the difference between the combined vibration waveform signal on the spindle side shown in FIG. 8 and the separated vibration waveform signal after the inverse fast Fourier transform processing of the separated vibration waveform spectrum signal. It is a graph, and (b) is a graph showing a differential vibration waveform signal obtained by taking a difference between the combined vibration waveform signal on the driven shaft side and the separated vibration waveform signal obtained by subjecting the separated vibration waveform spectrum signal to inverse fast Fourier transform processing. It is.
In FIG. 9A, the difference vibration is obtained by taking the difference between the combined vibration waveform signal 15a on the main shaft 30 side and the separated vibration waveform signal 19a (not shown) of the main shaft side sensor 35 after zero processing by the waveform separation processing unit 6. Since the waveform signal is obtained, only the abnormal residual component on the driven shaft 31 side appears. The reason is that in the separated vibration waveform signal 19a, only the abnormal component on the driven shaft 31 side is deleted, and only the abnormal component on the main shaft 30 side is on, so both the main shaft 30 side and the driven shaft 31 side overlap. When the difference from the combined vibration waveform signal 15a is taken, only the abnormal residual component on the counterpart driven shaft 31 side appears. Similarly, (b) takes the difference between the combined vibration waveform signal 15b on the driven shaft 31 side and the separated vibration waveform signal 19b (not shown), so that only the abnormal residual component on the main shaft 30 side appears.
In this embodiment, since the combined vibration waveform signals 15a and 15b when an abnormality has occurred are selected as the objects of the difference, the main shaft side sensor 35 shows an abnormal residual component on the driven shaft 31 side, and is driven. In the shaft-side sensor 36, an abnormal residual component on the main shaft 30 side appears. As described in the embodiment, if the normal-time vibration waveform signals 20a and 20b (not shown) are selected, either Are not superimposed, it is now possible to detect abnormal residual components from the components of the closest dynamic equipment, that is, differential vibration waveform signals 21a and 21b (not shown). .
In any case, since the residual is detected, information on the abnormal vibration waveform is obtained, and the difference signal threshold values 22a and 22b (not shown) described in the embodiment are used for those intensities. Thus, it is possible to evaluate the presence and degree of abnormality. The evaluation result 23 can be obtained and output as the evaluation result.

FIGS. 10A and 10B are graphs in which the differential vibration waveform signals shown in FIGS. 9A and 9B are fast Fourier transformed again and displayed as power spectra, respectively. According to FIG. 10A, it is understood that 7.75 Hz, 15.44 Hz, 23.19 Hz, and 30.64 Hz, which are vibration waveform components on the driven shaft 31 side, are detected as the output from the main shaft side sensor 35. it can. Further, in (b), it can be understood that 8.33 Hz, 16.88 Hz, 24.93 Hz, and 33.24 Hz, which are vibration waveform components on the main shaft 30 side, are detected as the output from the driven shaft side sensor 36. That is, it can be understood that separation can be performed with extremely high accuracy even in an environment where adjacent frequencies are mixed.
These evaluations and their outputs have already been described with reference to FIGS. 9 (a) and 9 (b).
In the present embodiment, the signals and threshold values obtained are not particularly described for operations for readable storage in the database and for reading from the database, but in actuality, in the embodiment. Implement as described.

  The invention described in claims 1 to 9 can be applied as a state monitoring system for a dynamic facility such as a manufacturing / processing plant or a power plant.

  DESCRIPTION OF SYMBOLS 1 ... Dynamic equipment state monitoring system 2 ... Sensor 3 ... Vibration frequency measurement part 4 ... Envelope processing part 5 ... Fast Fourier transform processing part 6 ... Waveform separation processing part 7 ... Inverse fast Fourier transform processing part 8 ... Residual calculation part DESCRIPTION OF SYMBOLS 9 ... Evaluation part 10 ... Output part 11 ... Synthetic vibration waveform database 12 ... Spectrum database 13 ... Separation wave waveform database 14 ... Evaluation database 15, 15a, 15b ... Synthetic vibration waveform signal 16, 16a, 16b ... Synthetic wave waveform after envelope processing Signals 17, 17a, 17b ... Synthetic vibration waveform spectrum signals 18, 18a, 18b ... Separate vibration waveform spectrum signals 19, 19a, 19b ... Separate vibration waveform signals 20, 20a, 20b ... Normal vibration waveform signals 21, 21a, 21b ... Differential vibration waveform signal 22, 22a, 22b ... Value 23 ... Evaluation result 24 ... Composite vibration waveform power spectrum signal 30 ... Main shaft 31 ... Driven shaft 32 ... Main shaft side gear 33 ... Driven shaft side gear 34 ... Rotor 35 ... Main shaft side sensor 36 ... Driven shaft side sensor X, Y ... Zero processing frequency

Claims (9)

In an environment where there are n (n: an integer of 2 or more) vibration sources, a dynamic facility state monitoring system capable of separating a composite vibration waveform generated from the n vibration sources,
N sensors capable of receiving a composite vibration wave generated from the n vibration sources;
A frequency measuring unit for processing n synthesized vibration waves received by the n sensors to obtain n synthesized vibration waveform signals;
A fast Fourier transform processing unit that obtains a spectrum signal by performing fast Fourier transform (hereinafter, sometimes referred to as FFT) on the n synthesized vibration waveform signals obtained by the frequency measuring unit;
For the n spectrum signals obtained by the fast Fourier transform processing unit, a difference in intensity is taken for each spectrum, and only a spectrum signal of a spectrum in which the difference is positive with respect to any other spectrum signal. The waveform separation processing unit that separates only the vibration wave component generated from the closest vibration source among the n vibration sources, with the spectral signal of the other spectrum being set to 0,
An inverse fast Fourier transform processing unit that obtains a vibration waveform signal from the closest vibration source by performing an inverse fast Fourier transform process on the spectrum signal separated by the waveform separation processing unit;
A residual calculation unit that calculates a difference between a vibration waveform signal at a normal time from the closest vibration source measured in advance and a vibration waveform signal from the closest vibration source obtained from the inverse fast Fourier transform processing unit; ,
An evaluation unit that compares the difference with a desired threshold value and determines that the closest vibration source is not normal when the deviation is larger than the threshold value;
A state monitoring system for dynamic equipment, comprising:   An envelope processing unit that envelopes n synthetic vibration waveform signals obtained by the frequency measurement unit is provided, and the fast Fourier transform processing unit inputs a post-envelope synthetic vibration waveform signal generated by the envelope processing unit The dynamic equipment state monitoring system according to claim 1, wherein a spectrum signal is obtained by performing a fast Fourier transform process.   The synthesized vibration wave generated from the n vibration sources is a synthesized sound wave, and instead of n sensors capable of receiving the synthesized vibration wave, the direction is changed n times with a single directivity and the n is changed. A microphone capable of measuring a synthesized sound wave generated from each vibration source, and the frequency measuring unit processes the synthesized sound wave received by the microphone to obtain n synthesized vibration waveform signals as n synthesized vibration waveform signals. The dynamic equipment state monitoring system according to claim 1, wherein a signal is obtained. In an environment where n (n is an integer of 2 or more) vibration sources are present, a state monitoring method for a dynamic facility capable of separating a composite vibration waveform generated from the n vibration sources,
Receiving a combined vibration wave generated from the n vibration sources using n sensors;
A frequency measurement step of processing each of the combined vibration waves received in the receiving step to obtain n number of combined vibration waveform signals;
A fast Fourier transform processing step of obtaining a spectrum signal by performing a fast Fourier transform on the n synthesized vibration waveform signals obtained in the frequency measurement step;
For the n spectral signals obtained in the fast Fourier transform processing step, a difference in intensity is taken for each spectrum, and only a spectral signal of a spectrum in which the difference is positive with respect to any other spectral signal. The waveform separation processing step of separating only the vibration wave component generated from the nearest vibration source among the n vibration sources, with the spectral signal of the other spectrum being set to 0,
An inverse fast Fourier transform processing step for obtaining a vibration waveform signal from the closest vibration source by performing an inverse fast Fourier transform process on the spectrum signal separated in the waveform separation processing step;
A residual calculation step of calculating a difference between a vibration waveform signal in a normal state from the closest vibration source measured in advance and a vibration waveform signal from the closest vibration source obtained in the inverse fast Fourier transform processing step; An evaluation step of comparing the difference with a desired threshold value and determining that the closest vibration source is not normal when the deviation is larger than the threshold value;
A state monitoring method for dynamic equipment, comprising:   An envelope processing step for performing envelope processing on the n synthetic vibration waveform signals obtained in the frequency measurement step, and the fast Fourier transform processing step inputs the post-envelope synthetic vibration waveform signal generated in the envelope processing step 5. The dynamic equipment state monitoring method according to claim 4, wherein a spectrum signal is obtained by performing a fast Fourier transform process.   The synthesized vibration wave generated from the n vibration sources is a synthesized sound wave, and the reception step changes the direction n times with a single directivity instead of the n sensors, and the n number of the vibration sources are changed. A reception step of measuring a synthetic sound wave generated from a vibration source using a microphone, wherein the frequency measurement step processes each of the synthetic sound waves received by the microphone to obtain n synthetic vibration waveform signals. 6. The dynamic equipment state monitoring method according to claim 4, wherein a synthesized sound waveform signal is obtained. A dynamic facility condition monitoring program executed by a computer to separate a composite vibration waveform generated from n vibration sources,
Receiving a combined vibration wave generated from the n vibration sources using n sensors;
A frequency measurement step of processing each of the combined vibration waves received in the receiving step to obtain n number of combined vibration waveform signals;
A fast Fourier transform processing step of obtaining a spectrum signal by performing a fast Fourier transform on the n synthesized vibration waveform signals obtained in the frequency measurement step;
For the n spectral signals obtained in the fast Fourier transform processing step, a difference in intensity is taken for each spectrum, and only a spectral signal of a spectrum in which the difference is positive with respect to any other spectral signal. The waveform separation processing step of separating only the vibration wave component generated from the nearest vibration source among the n vibration sources, with the spectral signal of the other spectrum being set to 0,
An inverse fast Fourier transform processing step for obtaining a vibration waveform signal from the closest vibration source by performing an inverse fast Fourier transform process on the spectrum signal separated in the waveform separation processing step;
A residual calculation step of calculating a difference between a vibration waveform signal in a normal state from the closest vibration source measured in advance and a vibration waveform signal from the closest vibration source obtained in the inverse fast Fourier transform processing step; An evaluation step of comparing the difference with a desired threshold value and determining that the closest vibration source is not normal when the deviation is larger than the threshold value;
A condition monitoring program for dynamic equipment, characterized in that   An envelope processing step for performing envelope processing on the n synthetic vibration waveform signals obtained in the frequency measurement step, and the fast Fourier transform processing step inputs the post-envelope synthetic vibration waveform signal generated in the envelope processing step The dynamic facility state monitoring program according to claim 7, wherein a spectrum signal is obtained by performing a fast Fourier transform process.   The synthesized vibration wave generated from the n vibration sources is a synthesized sound wave, and the reception step changes the direction n times with a single directivity instead of the n sensors, and the n number of the vibration sources are changed. A reception step of measuring a synthetic sound wave generated from a vibration source using a microphone, wherein the frequency measurement step processes each of the synthetic sound waves received by the microphone to obtain n synthetic vibration waveform signals. 9. The dynamic facility state monitoring program according to claim 7, wherein a synthesized sound waveform signal is obtained.